James Seals asked in the comments to Places to go and people to meet: The One Spot series from Moebius Adventures (responding to my comments about magic shops),


Can I ask – what do you do when your players want to sell magic weapons? In the past I have just allowed them to be sold to Ye Olde Magic Item Shoppe for 50% of the DMG value, but I totally agree with your points above and would prefer a better solution.


The short answer is that it depends on the campaign, which is not very helpful. So let me offer an answer that’s a little more robust…

Defining The Problem: The Longsword Economy

Let’s start by sketching in a little background. I’ll assume that the rather silly coinage given in D&D/Pathfinder is correct.

It took the master swordsmith a year to make the three swords featured in The Lord Of The Rings – with the best advantages modern technology could offer his very traditional craft. Yet, those are the sort of weapons most people think of when they think about Longswords in RPGs.

To make any more than that, you have to cut down on the quality of the workmanship. Which is fine, there’s a lot of room to maneuver. One step removed from weapons of that quality you would get the weapons provided to elite forces within the army – dress swords and the like. Assume that you could make 4 times as many of these 2nd-tier weapons in a year, and that this is what is priced in the Core Rules as a standard masterwork sword. That gives us 12 a year, a convenient number to work with.

One step down in quality from that again, a further 4-fold increase in production, gives us the sort of practical weapon that would have been issued to members of the army of the kingdom. Each of these takes about a week to make, another convenient number.

Another step down in quality produces poor weapons – so poor that no-one would want them. And one step further removed in quality produces the sort of weapons we see being forged in Isengard in The Lord Of The Rings for the Orcs. Low quality, no decoration, a single bladed edge, and something approaching mass production. So those can be knocked out at the rate of about 16 a week – 1 day to prepare the moulds etc, 1 day to cast them, and sharpening 4 a day – per smith.

If we employ the price of a Longsword as the standard of a fantasy economy this is enough to define an average annual income. This is convenient because the price of a longsword is what I use to determine conversion rates for “…And A 10-foot Pole”, as I explained in How Much Is That Warhorse In The Window? – Pricing Of Goods in D&D.

So the average Blacksmith has an income, under these assumptions, of about 50 times the price of a longsword. His disposable income will be a lot smaller – cost of materials & tools, subsidizing apprentices, rents, food, lodgings, taxes, tithes, and donations, saving for dowries, what-have-you. With so many demands on his purse, it would seem unbelievable for more than 5% of his income to be left at the end of the day, and I suspect that 1% or less would be closer to the mark. At 15gp (3.x/Pathfinder) for a Longsword, that gives a disposable income of 15 x 50 x 0.05 = 37.5gp, per annum.

I worked this problem from a different angle two years ago in a sidebar within Loot As Part Of The Plot: Making, Earning, Finding, Analyzing, Using, Selling, and Destroying Loot, reproduced below, and came up with a figure of cash-on-hand of 240gp. (I got a bit carried away and quoted the entire “Selling” section from that article, because it’s relevant).

Fantasy Economic Assumptions: A Venting

<begin venting>

This should be a lot harder than most GMs make it. “I have a 10,000gp gem that I’d like to trade in for gold pieces”. “I have a +2 dagger to sell.” “How much will you give me for a Sphere Of Annihilation?”

How many NPCs will have 10,000gp on hand? Of those that do, how many are willing to tie it all up in a single valuable? What is that money supposed to be used for? Who will object to it being used in this way? Who will object to the PCs having such a valuable and wish to redistribute the wealth? Can anyone else lay legal claim to it? Is there a legal requirement to make change when claiming payment for goods or services? (you would be astonished to learn how many countries have no such law – just the tradition of doing so. It is taken for granted…

For every seller, there has to be a buyer. And one of the first questions a GM should ask is “why” does this NPC want to buy the loot? How much is he willing to pay? What expenses will he incur? How much can he expect to make on the deal?

Fantasy economics generally has a number of holes in it in this department. In modern times, a typical business has a profit, annually, before tax, of about 10-20% of its turnover – call it 15% for convenience. In theory, that equates to its markup, or profit margin, on the products and services that it provides; in practice, there are variables that this doesn’t take into account. And a successful business will have 5 years of profits – after taxes – as a cash reserve. So, a suit of full plate costs 1500gp according to the PHB; and it might take a skilled armorer a month to make such a suit. That gives 1500gp x 12 = 18000gp a year income. Ten percent of which is 1800 gp. Apply a modern maximum tax rate of, say, 50%, and you get 900gp per year. Five years at 900 gp gives the NPC a cash reserve of 4500 gp – applying modern standards.

How about less generous standards from a bygone era that is more directly comparable to the game setting? Profit margin: 30%, but 1/3 of the production (perhaps more, especially in times of war – and when isn’t a medieval society at war with someone?) goes to the Lord for free. Three years is a more appropriate cash reserve, because unexpected expenses are much higher and eat into the character’s money. And the tax rate is going to be more like 70-90% – call it 80%. Work out the numbers: 1500 x 3 = 4500, less the profit margin of 30%, means that the levee by the Lord costs 3150gp a year. 1500 x 9 = 13500, times 13% = 4050. Minus that 3150, leaves a net profit of 900gp a year. Take off taxes of 80% and we have net income of 180gp a year. Out of which the NPC has to buy food and pay rents and replace damaged tools and what have you – which might leave 130gp a year, being generous, or (more likely) 80gp. Three year’s reserve equals 240gp. That’s how much the armorer can afford to spend buying unwanted adventurer castoffs, no matter how much he might be able to eventually sell it for. He can’t afford to speculate; the people who might want to buy it from him might take ten years or more to come up with cash (Nobles and governments are notoriously poor at keeping accounts current).

</end venting>

Rather than requiring an economic analysis of every prospective purchase by the GM, there is a simpler answer: the NPCs have enough coin on hand to meet the GM’s story needs – no more, and no less.

What would more normally occur is this: The Blacksmith would offer to approach various people on behalf of the prospective seller, at a price of 50gp a day (1500gp divided by a month, neatly rounded), paid in advance, as an introduction fee; if the visit results in a sale, he would get a commission of 5% or perhaps 10% from the deal. He would put a cap on how much time he risked that was equal to half his gold reserve divided by 50gp a day – so a reserve of 250gp would permit him to spend two-and-a-half days trying to sell the armor. Anything more than that risks his livelihood. If the prospects were good, he might go as high as three or four days.

Throw in bureaucracy and red tape and travel time, and he will be doing well to approach more than two prospective customers in that time frame. If they aren’t interested, neither is he.

There is one point in the text of Assassin’s Amulet that Johnn, after reading it, said completely changed his views on game economics in at least one respect. I pointed out, in the section on the price of an assassination contract, that whatever the fee charged was, every assassination required someone to have paid that fee. Which means they had that much money on hand to expend on the assassination, and were willing to expend it – that was how much it was worth to them. If their motives were profit-related, they had to expect to make at least that much more money from the deal in the long run.

The same applies to every purchase of an item from a PC. The character doing the buying must have that much on hand, and owning the item in question has to be worth their investing all of it in the item. How do they have that much money? Why do they want the item badly enough to buy it?

Selling a magic item – or a rare gem – or a work of art – should immediately raise serious questions in the mind of a PC. If it doesn’t, it’s a sure bet that the GM has been neglecting this type of plot hook.

Regardless of which value you choose – 37.5gp or 80gp per year – the typical NPC blacksmith is not going to be able to afford to buy even a +1 weapon very often.

The Distribution Of Incomes

But wait, that’s just defining the average wage for a working man. Some professionals will earn more, and some will earn less.

There are two types of curve that can fit earning levels; a dumbbell, and a non-dumbbell. We can simulate the first by adding dice such that the average gives us that average wage.

Dumbbell: 3d20 + 2d6 – 1 gives a range of 4 to 72 and an average of 37.5. This means that some professionals will earn 1.92 times that average – which isn’t all that much, and doesn’t seem all that realistic. There are other solutions possible, but they all yield a maximum that isn’t that much removed from double the average.

The alternative is a weighted distribution which assigns a low probability of a much higher value. d20 x d% / (2d12 + (2d6 x 2d6 / (d12 x d12))) gives an average of 37.44gp – but a peak of 993.1gp. The probability distribution looks like this:

That’s incredibly messy, in terms of a result. But we’re not actually interested in generating a die roll that yields these results; we can get there by a much simpler road: ratios.

If we say that 1 in five earns three times as much as the average (instead of simply saying half the population earns more than the average), we can extrapolate out to our heart’s content:

  • Base: 3125 earning an average 37.5gp;
  • pass 1: 625 earning an average 112.5gp;
  • pass 2: 125 earning an average 337.5gp;
  • pass 3: 25 earning an average 1012.5gp;
  • pass 4: 5 earning an average 3037.5gp;
  • pass 5: 1 earning an average 9112.5gp;

All we then have to do is increase the numbers of people earning substantially less than the average to compensate for the uber-wealthy so that our average is maintained:

  • 625 x 112.5 = 68850;
  • 125 x 337.5 = 42187.5; + 68850 = 111037.5;
  • 25 x 1012.5 = 25312.5; + 111037.5 = 136350;
  • 5 x 3037.5 = 15187.5; + 136350 = 151537.5;
  • 1 x 9112.5 = 9112.5; + 151537.5 = 160650;
  • 625+125+25+5+1 = 781;

so N people earning 10gp average per annum have to add to the 781 earning a total of 160,650gp to give our 37.5 average:

(160,650 + N x 10) / (N + 781) = 37.5;

rearranging and expanding gives: 27.5xN = 160,650 – (37.5 x 781);

…which calculates out as N=4776.8. So we can add another line to the top of our table, reading:

  • 4777 earning an average 10gp;

…and the grand total becomes 3125+4777+871=8773 people.

Now, the real uber-wealthy can afford something better than a second-hand +1 weapon. But there’s going to be a middle ground.

The Income Of Adventures

First-level adventures typically yield somewhere between 100 and 1000gp. Divided four or five ways, that gives something like 20-200gp per party member. The average yield would be somewhere close to the middle of that range, so call it 100gp for convenience. It can take anywhere from a couple of days, game time, to a few months (including rest & recovery, afterwards), so there can be anywhere from 3 or 4 to 150 of them in a year, but the average trend would be toward the lower end of that range – maybe 10-20 would be median. Call it 15 for the sake of convenience.

That’s an income of 1500gp per annum. It only skyrockets with increasing character levels.

PC expenses are woefully out of line in comparison. Three-to-five GP a week is ample for a lavish scale of living. Leveled characters are automatically, under standard D&D/Pathfinder canon, going to be amongst the wealthiest people in the town/city/kingdom whenever they rock up somewhere. Let’s be even more extravagant and set expenses at 500gp per annum.

That still leaves a disposable income of 1000gp per year. Our blacksmith – who is assumed to earn an average wage for a professional – has to work for twelve-and-a-half years to twenty-six-plus years to get that much money on tap.

It was this line of thought that led me to shift Fumanor (during campaign creation) to a silver standard, and to insert a new currency, bronze pieces, between copper and silver pieces. Assuming that when a published module says “gp” it means “sp” applies the currency conversion (10 sp = 1 gp) to the annual income of an adventurer and makes this a profession that pays a little better than a skilled blacksmith earns – which is reasonable, given the relative risks.

How many adventurers?

In The Foundation Of Averages: Psychohistory and RPG Rules I looked at populations in terms of character levels. At the time, I was looking at the number of levels earned by the time the adventurers retired, and showed that for every character with high levels, an absolutely ridiculous number of first-level characters were statistically required. Assuming that only one in five adventurers survive and stay active long enough to reach their next character level, for every 20th level character there had to be 19,073,486,328,125 first level characters. 19 million million 1st level adventurers! (I’ve avoided using the term “Billion” because it means different things in different countries). At a more modest one-in-two progression rate, it works out that for every million adventurers, 1 will be 20th level. Here’s the full breakdown:

  • 20th level: 1
  • 19th level: 2
  • 18th level: 4
  • 17th level: 8
  • 16th level: 16
  • 15th level: 32
  • 14th level: 64
  • 13th level: 128
  • 12th level: 256
  • 11th level: 512
  • 10th level: 1,024
  •  9th level: 2,048
  •  8th level: 4,096
  •  7th level: 8,192
  •  6th level: 16,384
  •  5th level: 32,768
  •  4th level: 65,536
  •  3rd level: 131,072
  •  2nd level: 262,144
  •  1st level: 524,288

…for a total of 1,048,575 adventurers. This was the highest ratio that I thought made it reasonable for PCs of any given level to encounter an enemy of equivalent level.

But these levels are extraordinarily sensitive to the ratio, because it’s applied as an exponential factor. If, for example, the correct ratio was 1.95 instead of 1 in two, we get:

  • 20th level: 1
  • 19th level: 1.95
  • 18th level: 3.8025
  • 17th level: 7.41488
  • 16th level: 14.459
  • 15th level: 28.195
  • 14th level: 54.9803
  • 13th level: 107.212
  • 12th level: 209.063
  • 11th level: 407.673
  • 10th level: 794.962
  •  9th level: 1,550.18
  •  8th level: 3,022.85
  •  7th level: 5,894.56
  •  6th level: 11,494.4
  •  5th level: 22,414.1
  •  4th level: 43,707.5
  •  3rd level: 85,229.6
  •  2nd level: 166,198
  •  1st level: 324,086

… for a total of 665,228 adventurers for every 20th level character. We lose 200,000 first-level adventurers alone! And a ratio of 1.75 gives:

  • 20th level: 1
  • 19th level: 1.75
  • 18th level: 3.0625
  • 17th level: 5.35938
  • 16th level: 9.37891
  • 15th level: 16.4131
  • 14th level: 28.7229
  • 13th level: 50.2651
  • 12th level: 87.9639
  • 11th level: 153.937
  • 10th level: 269.39
  •  9th level: 471.433
  •  8th level: 825.008
  •  7th level: 1,443.76
  •  6th level: 2,526.58
  •  5th level: 4,421.52
  •  4th level: 7,737.66
  •  3rd level: 13,540.9
  •  2nd level: 23,696.6
  •  1st level: 41,469.1

and a total of 96,759.8 characters for every 20th level character. or, to put it another way, for every 1,048,575 adventurers – the total we got per 20th level character with a 1-in-2 ratio – we get about 10 additional 20th level characters by boosting the survival rate to a ratio of 1-in-1.75.

Let’s take that 1.75-ratio table and multiply by the character wealth by level according to Pathfinder (in thousands of gp, and with a value of 350gp for 1st level because they don’t include one):


  • 20th level: 880
  • 19th level: 1,198.75
  • 18th level: 1,623.12
  • 17th level: 2,197.35
  • 16th level: 2,954.36
  • 15th level: 3,939.14
  • 14th level: 5,313.74
  • 13th level: 7,037.11
  • 12th level: 9,500.1
  • 11th level: 12,622.8
  • 10th level: 16,702.2
  •  9th level: 21,685.9
  •  8th level: 27,225.3
  •  7th level: 33,928.4
  •  6th level: 40,425.3
  •  5th level: 46,426
  •  4th level: 46,426
  •  3rd level: 40,622.7
  •  2nd level: 23,696.6
  •  1st level: 14,514.2

…for a total economic worth of 358,919,000 gp – or an average of 358919000/96759.8 = 3709.38gp each. (I hinted at these sort of results in the previously mentioned article. They are even more extreme with higher survival ratios, with the peak point shifting to lower character levels. The closer to 1-1 the ratio gets, the more the peak shifts toward the middle of the character level range, and the more the average wealth declines per adventurer. At a ratio of 1-in-2.75, the peak is 3rd level, just barely in front of second level, the average wealth per adventurer is down to 1141.33gp each, and the total economic wealth of adventurers as a group is the utterly preposterous 77,874,300,000gp).

The Preposterous Placement of Treasures

Okay, so let’s assume that we have 665,000-odd adventurers in a kingdom, in parties averaging 4.5 members (some have 4, some 5, some 3, some 6, and so on), and that each party of 4th-level characters has at least one adventure a year in which they recover an unwanted +1 item. According to the numbers above, that’s 7,738 adventurers, or about 1720 parties, or 1720 unwanted +1 swords, each worth 2000gp according to Pathfinder.

Three-point-four-four million gp worth.

Who can afford them?

There are two obvious answers: Other adventurers and the uber-wealthy.

But it’s not that simple.

The Uber-wealthy as buyers

Our extrapolation of income distribution gave us 6 people in 8773 who could afford to buy a +1 weapon, once a year. Every time we increase the number of income earners that we’re talking about by a factor of 5 (because we said one-in-five in those extrapolations), we would need to add another line to our table and recalculate the number of low-income earners accordingly.

6 to 30 to 150 to 750 to 3,750. That’s four more lines to our income distribution table, and a base income-earning population of 3125x5x5x5x5, or 1,953,125 people working as professionals for wages and not growing food or something else along those lines, or digging up raw materials, or whatever.

But most of the people on the resulting list will be able to afford something better. A +2 weapon only costs, new, four times as much as a +1 weapon. The slice of the market who can both afford to buy a +1 weapon, can’t afford better (including buying a new +1 weapon with their own crest or mark on it), and is willing to buy a second-hand +1 weapon, is not going to go up at anywhere near the rate of increase – or to put it another way, the further we extend the table, the smaller the percentage of the whole they will form.

No way, no how – there simply aren’t going to be enough customers amongst the uber-wealthy to accommodate 1720 unwanted +1 weapons coming on to the market each and every year – not with our current set of assumptions.

How many of these things got made, anyway? And how many are already in circulation?

But, with the bases covered, I’m at least ready to get back to the original question.

The Utopian Answer:

The price of goods is based on supply vs. demand. If there’s a lot of demand and not a lot of supply, the price will go up. If there’s a lot of supply and not a lot of demand, no matter what the notional value may be, it’s actual value – what you will get when you sell it, will be coppers on the silver or worse.

So, plausible self-consistent answer number one is:

  • Basic Magic items are easy and cheap to produce, and are worth about 1/10th of the price listed;
  • There is plenty of raw materials and food, and a high standard of living;
  • Everyone and his brother – if they are even moderately prosperous – can afford a +1 weapon;
  • There is a thriving professional class, supported by this cornucopia;
  • Adventuring is a common practice, viewed as a bit of a lark, something like taking a gap year;
  • lethality in the campaign is low;
  • Treasures and Hordes should be at least as large as those officially listed if not higher;
  • The gp is the basic unit of currency, and there is a ready market for anything that can be found;
  • ‘Magic shops’ exist, but what’s available depends on what’s been sold to them. It’s buyer’s luck.

The Dystopian Answer:

  • Magic is rare, and even Basic Magic items are difficult and expensive to make (if it’s possible at all), and the prices quoted in gp in the official rules are correct;
  • There is a limit to how much raw material and food there are, and beggars and paupers are common;
  • only a few can afford to buy even a +1 weapon;
  • There are few professionals, but lots of low-skill craftsmen;
  • Adventuring is a rare career path; but can be exceptionally lucrative, if you survive;
  • The sp is the basic unit of currency; all income levels should be divided by 10, prices of basic commodities likewise;
  • Magic items are worth the prices printed in the books, but should be nowhere near as common as shown in most modules and treasure tables.

My Choice – Answering James’ Question

I’ve gone with the first choice in past campaigns, most recently Rings Of Time (which started with the PCs inheriting a Dragon’s Hoard worth 353,000gp, captured by higher-level party members, all of whom died in the attempt).

I’ve also used a hybrid showing the transition from one to the other in a previous campaign. But, for the most part, the campaigns I have run have focused on the Dystopian orientation, simply because it makes for more interesting campaigns.

PCs in Fumanor have gotten into epic levels before finding a +2 weapon (though there were some missed opportunities along the way, had they realized). And it’s even rarer to find a weapon that doesn’t have a drawback or limitation of some sort attached to it.

But, at the same time, there are some unique magic items in somewhat wider circulation, like wax seals that can be attached to any weapon and will temporarily yield a +1 benefit – then removed, ready to be attached to another. Healing potions are ubiquitous, but are specific to the species for whom they were brewed – and have strange side effects when consumed by members of other races. There are places that are blessed, and confer bonuses to those who fight there if they are acting in the name of Good, and others that are cursed and do the opposite. There are places where bonuses that do not normally stack can do so, and places where all bonuses are not stackable. There are places where members of one specific race or class receive a bonus, or a penalty, and places where everyone not of a given race is affected.

In Shards Of Divinity, Magic is becoming unreliable, and the world is transiting from semi-utopian to dystopian. Magic items are rare, but more common than in Fumanor, for the moment, though no-one can create new ones of any great power any more; only those recovered from ancient times exist, and many of them are far more powerful than a mere +1. At the current time, any spell has a 3-in-20 chance of failing when cast, but magic items are unaffected. In the near future, that will progress to a 1-in-4, then a 1-in-5, and so on, and magic items will begin to be affected as well. The more highly magical an item, the lower its chance of failing. One of the PCs primary goals is doing something about this situation; they are the only people in existence who know why it is happening in the first place. But right now, magic items are more reliable than the magic used to create them, and so their value is artificially inflated by demand to the levels indicated by the rulebooks.

And, of course, your third option is to do as I suggested in the quoted section above – treat attempts to sell magic items as though they were attempts to sell a previously-undiscovered Rembrandt or the Star Of India, and an oportunity to throw a different plot at the PCs.

In other words, I have no one answer; I have, rather, a set of considerations starting with the way the economy works in the game world, which in turn is tied to the sociology and a host of other factors, which I use to develop an internally-consistent answer that is unique to that particular campaign.

On a completely unrelated topic:

I don’t know about any of our other subscribers, but Monday’s article didn’t lob into my inbox the way it should have. It’s possible that this is related the extraordinary size of the article, or that something broke in our last systems update. Until I see what happens with this article, I’m going to assume that it was a size-related problem, in which case, here’s a link to the article that aparrantly did not get sent: A Blogdex Celebration: Campaign Mastery’s (official) 500th post!

Related Posts with Thumbnails
Print Friendly